Deicing device pole-to-ground fault analysis method and system considering transient characteristics

Abstract

The invention discloses an ice melting device pole-to-ground fault analysis method and system considering transient characteristics, and relates to the technical field of power system fault analysis, and the method comprises the steps: building a simulation environment according to a topological structure and historical sub-module operation data of a to-be-tested symmetric single-pole ice melting device; performing fault simulation on the to-be-tested symmetric single-pole ice melting device according to the pole-to-ground fault parameters to obtain voltage transient waveforms reflecting sub-modules in different switching states; performing voltage stress analysis on the voltage transient waveform to obtain voltage stress distribution data and waveform abnormal points of the sub-modules; inputting the voltage stress distribution data and the waveform abnormal points into a pre-trained fault prediction model for processing to obtain fault prediction results corresponding to the sub-modules; and according to the fault prediction result, determining the fault risk level of the to-be-tested symmetric single-pole ice melting device, and improving the reliable operation of the symmetric single-pole ice melting device.

Description

Technical Field

[0001] The present invention relates to the technical field of power system fault analysis, and particularly to a method and system for analyzing the pole-to-ground fault of an ice melting device considering transient characteristics. Background Art

[0002] Research in the field of power systems is of crucial significance for ensuring the safe and stable operation of the power grid. Among them, the DC ice melting technology, as a key means to cope with ice and snow disasters, is directly related to the reliability of transmission lines and the safety of social economy. With the intensification of climate change, the threat of ice disasters to the power system is becoming increasingly severe. The symmetric monopolar ice melting device has become a research hotspot due to its high efficiency and flexibility. It realizes efficient ice melting through a modular multilevel converter (MMC). However, this structure also increases the probability of faults to a certain extent. For example, in the initial stage of the fault occurrence of the symmetric monopolar ice melting device, some arm sub-modules may be damaged due to excessive voltage stress, which will further lead to the instability of the power system.

[0003] Traditional fault analysis methods only perform simple steady-state analysis and cannot comprehensively capture the dynamic response characteristics during the fault process, resulting in poor accuracy in fault risk analysis and prediction. Reasonable and reliable control strategies or maintenance measures cannot be formulated, thereby affecting the ability of the ice melting device to cope with faults.

[0004] Therefore, how to effectively analyze the pole-to-ground fault of the symmetric monopolar ice melting device and improve the fault response ability of the ice melting device has become a technical problem that needs to be urgently solved by those skilled in the art. Summary of the Invention

[0005] The present invention provides a method and system for analyzing the pole-to-ground fault of an ice melting device considering transient characteristics, aiming to solve the problem of how to effectively warn and control risks by deeply studying the transient distribution characteristics of the voltage stress of the arm sub-modules during the pole-to-ground fault of the symmetric monopolar ice melting device, and maintain the reliability and safety of the power system.

[0006] To solve the above technical problems, an embodiment of the present invention provides a method for analyzing the pole-to-ground fault of an ice melting device considering transient characteristics, including: Create a simulation environment according to the topological structure and historical sub-module operation data of the symmetric monopolar ice melting device to be measured; Perform a fault simulation on the symmetric monopolar ice melting device to be measured with preset pole-to-ground fault parameters, and obtain the voltage transient waveforms of the sub-modules of the symmetric monopolar ice melting device to be measured under different switching states; Perform voltage stress analysis on the voltage transient waveforms to obtain the voltage stress distribution data and waveform abnormal points of the sub-modules; Input the voltage stress distribution data and the waveform anomaly points into a pre-trained fault prediction model for processing to obtain the fault prediction results corresponding to the sub-modules; Determine the fault risk level of the to-be-tested symmetric monopolar ice melting device according to the fault prediction results.

[0007] Further, before performing fault simulation on the to-be-tested symmetric monopolar ice melting device with preset pole-to-ground fault parameters, it further includes: Extract features from the historical switching data in the historical sub-module operation data to construct a switching feature data set; Input the switching feature data set into a pre-constructed statistical distribution model, and output the feature distribution during the sub-module switching process; Adjust the sub-module parameters of the to-be-tested symmetric monopolar ice melting device according to the feature distribution, and update the simulation environment with the adjusted sub-module parameters and the topological structure.

[0008] Further, the voltage stress analysis of the voltage transient waveform to obtain the voltage stress distribution data and waveform anomaly points of the sub-module includes: Extract the voltage signal data in the voltage transient waveform, and divide the input and cut-off states of the sub-module according to the voltage signal data; According to the division results, calculate the voltage stress index values in each switching state respectively; Perform statistical analysis on the voltage stress index values to obtain the voltage stress distribution data, and mark the waveform anomaly points in the voltage transient waveform according to the screening factor calculated from the voltage stress index values.

[0009] Further, the performing statistical analysis on the voltage stress index values to obtain the voltage stress distribution data, and marking the waveform anomaly points in the voltage transient waveform according to the screening factor calculated from the voltage stress index values includes: Perform multi-index joint analysis on the voltage stress index values to fit the voltage stress distribution curve of the sub-module; and, Calculate the overvoltage multiple and the voltage change rate according to the voltage stress index values, and use the overvoltage multiple and the voltage change rate as double screening factors; When any index in the double screening factors exceeds the preset anomaly threshold, mark the waveform anomaly points in the voltage transient waveform.

[0010] Further, the dividing the input and cut-off states of the sub-module according to the voltage signal data includes: Calculate the difference between the voltage signal data and the preset rated voltage value of the sub-module, and set the division threshold with the difference; Based on the division threshold, traverse the voltage transient waveform using the sliding window algorithm to determine the switching states of the sub-modules.

[0011] Furthermore, the process of performing voltage stress analysis on the voltage transient waveform further includes: Adjust the topological connection mode of the sub-modules in the symmetric single-pole ice melting device to be measured, and update the simulation environment with the adjusted first topological structure; In the updated simulation environment, control the input / removal logic of the sub-modules to re-perform fault simulation; Obtain the corresponding comparison voltage transient waveforms of the sub-modules under different topological structures according to the simulation results; Determine the comparison voltage stress distribution data according to the comparison voltage transient waveforms; Compare and analyze the comparison voltage stress distribution data and the voltage stress distribution data to determine the high-order indicators affecting the voltage stress of the sub-modules.

[0012] Furthermore, the process of comparing and analyzing the comparison voltage stress distribution data and the voltage stress distribution data to determine the high-order indicators affecting the voltage stress of the sub-modules includes: Preprocess the first topological structure, sub-module switching states, the voltage stress distribution data, and the comparison voltage stress distribution data to obtain the input dataset to be input; Input the input dataset to be input into the trained random forest model, and output the importance scores of the indicators affecting the voltage stress; Sort the importance scores, and determine the high-order indicators according to the sorting results.

[0013] Furthermore, the process of performing voltage stress analysis on the voltage transient waveform further includes: According to the voltage stress distribution data, compare the first voltage stress differences of the sub-modules in different switching states; According to the comparison voltage stress distribution data, compare the second voltage stress differences of the sub-modules under different topological connection modes; Input the preprocessed first voltage stress differences and the second voltage stress differences into the pre-trained difference classification model to obtain the stress difference levels of each sub-module; Correct the topological structure and switching state parameters of the sub-modules according to the stress difference levels, and re-iterate the simulation analysis process.

[0014] Furthermore, the process of determining the fault risk level of the symmetric single-pole ice melting device to be measured according to the fault prediction result includes: Extracting features from the fault prediction results to obtain a fault feature data set; The fault feature data set is input into a pre-trained risk prediction model to obtain the fault risk level, and a corresponding maintenance strategy for the symmetrical monopolar ice melting device to be tested is formulated based on the fault risk level.

[0015] Another embodiment of the present invention provides an ice melting device pole-to-ground fault analysis system considering transient characteristics, comprising: A simulation initialization module is used to create a simulation environment according to the topological structure of the symmetrical monopole ice melting device to be tested and the historical submodule operation data; A fault simulation module, used for performing fault simulation on the symmetrical monopole ice-melting device to be tested with preset pole-to-ground fault parameters, and obtaining voltage transient waveforms of submodules of the symmetrical monopole ice-melting device to be tested under different switching states; A transient analysis module, used to perform voltage stress analysis on the voltage transient waveform, and obtain voltage stress distribution data and waveform abnormal points of the submodules; A fault prediction module, used for inputting the voltage stress distribution data and the waveform abnormal points into a pre-trained fault prediction model for processing to obtain a fault prediction result corresponding to the submodule; The risk analysis module is used to determine the fault risk level of the symmetrical monopolar ice melting device to be tested according to the fault prediction result.

[0016] Compared with the prior art, the embodiments of the present invention have the following advantages: The embodiment of the present invention builds an electromagnetic transient simulation environment including a bridge arm submodule model, accurately captures the voltage transient data in the initial stage of the fault through simulation, and quantifies the key parameters of the submodule voltage stress, thereby improving the accuracy of the analysis of the dynamic response; it also comprehensively reveals the influence of the voltage stress distribution under the fault scenario by analyzing the influence of different switching states, and establishes a classification prediction model for voltage stress index parameters to effectively identify the fault type, location and submodule switching state; thereby, it is possible to accurately assess the risk of submodule damage, so as to formulate reasonable and reliable device prevention strategies or maintenance measures, thereby ensuring the stable operation of the symmetrical monopole ice melting device under extreme working conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a schematic flow chart of a method for analyzing a pole-to-ground fault of an ice melting device taking into account transient characteristics in one embodiment of the present invention; Figure 2 It is a schematic diagram of the structure of a pole-to-ground fault analysis system for an ice melting device considering transient characteristics in one embodiment of the present invention; Description of the Drawings: M1, simulation initialization module; M2, fault simulation module; M3, transient analysis module; M4, fault prediction module; M5, risk analysis module. Detailed Implementation Manner

[0018] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. The purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without making creative efforts shall fall within the protection scope of the present invention.

[0019] In the description of the present application, the terms "first", "second", "third", etc. are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first", "second", "third", etc. may explicitly or implicitly include one or more of such features. In the description of the present application, unless otherwise stated, the meaning of "a plurality" is two or more.

[0020] In the description of the present application, it should be noted that unless otherwise clearly defined and limited, the terms "installed", "connected", "connected" should be understood in a broad sense. For example, it may be a fixed connection, a detachable connection, or an integral connection; it may be a mechanical connection or an electrical connection; it may be directly connected, or indirectly connected through an intermediate medium, and it may be the communication inside two elements. The terms "vertical", "horizontal", "left", "right", "up", "down" and similar expressions used herein are only for the purpose of illustration, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus cannot be understood as a limitation to the present invention. The term "and / or" used herein includes any and all combinations of one or more of the related listed items. For those of ordinary skill in the art, the specific meanings of the above terms in the present application can be understood according to specific circumstances.

[0021] In the description of the present application, it should be noted that unless otherwise defined, all technical and scientific terms used in the present invention have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present invention belongs. The terms used in the specification of the present invention are only for the purpose of describing specific embodiments, and are not intended to limit the present invention. For those of ordinary skill in the art, the specific meanings of the above terms in the present application can be understood according to specific circumstances.

[0022] When a symmetrical monopole ice melting device experiences a pole-to-ground fault, the overvoltage stress borne by the arm sub-module during the electromagnetic transient process may cause equipment damage. Based on this, an embodiment of the present invention provides a method for analyzing the pole-to-ground fault of an ice melting device considering transient characteristics. Specifically, please refer to Figure 1 , Figure 1 which shows a schematic flowchart of the method for analyzing the pole-to-ground fault of an ice melting device considering transient characteristics in one embodiment of the present invention, including the following steps: S1. Create a simulation environment according to the topological structure and historical sub-module operation data of the symmetrical monopole ice melting device to be tested.

[0023] Obtain the topological structure of the symmetrical monopole ice melting device to be tested. The topological structure of the symmetrical monopole ice melting device is usually designed based on a modular multilevel converter, and the core is to realize the DC ice melting function through the series and parallel combination of sub-modules. Exemplarily, in this embodiment, a high-precision electromagnetic transient simulation tool (such as PSCAD / EMTDC or MATLAB / Simulink) can be selected to establish a simulation model including a sub-module cascade structure. Correspondingly, various parameters for the reasonable operation of the device should be configured in the simulation software, and the parameter configuration can be carried out through the obtained historical sub-module operation data, including configuring the capacitance value, rated voltage value, IGBT withstand voltage level, switching strategy (such as switching timing) of the arm sub-module, etc.

[0024] Exemplarily, the topology of the symmetrical monopole ice melting device to be tested in this embodiment can be that multiple arms are connected in parallel, and each arm is composed of several sub-modules connected in series. The sub-module can adopt a half-bridge or full-bridge form. For example, one arm of the device contains 10 sub-modules. If the rated voltage of each sub-module is 1 kV, the total voltage of the arm can reach 10 kV, thus meeting the DC output required for ice melting.

[0025] Furthermore, before the fault simulation, to ensure the simulation efficiency and accuracy, this embodiment will calibrate the constructed simulation model. Specifically, in this embodiment, the simulation parameters are optimized by analyzing the characteristic distribution of the sub-module during the historical switching process.

[0026] First, construct a feature dataset, extract features from the historical switching data in the historical sub-module operation data, construct a switching feature dataset, and input it into a pre-constructed statistical distribution model to output the characteristic distribution during the sub-module switching process. In this embodiment, the extracted switching features include switching frequency, the number of sub-modules involved in a single switching, date, and capacitor voltage value.

[0027] Subsequently, adjust the sub-module parameters of the symmetrical monopole ice melting device to be tested according to the characteristic distribution, and update the simulation environment with the adjusted sub-module parameters and topological structure.

[0028] Exemplarily, the switching operation records of the sub-module in the past year can be extracted from historical data. Suppose the record shows that the switching frequency is concentrated between 50Hz and 100Hz, and each switching involves 2 to 5 sub-modules. Using statistical analysis methods, characteristic distributions are obtained, such as maximum likelihood estimation, regression analysis, Poisson distribution, or normal distribution, etc. For example, the number of switching operations follows a normal distribution, with a mean of 3 sub-modules and a standard deviation of 1. This indicates high-frequency switching, and the capacitor needs to be charged and discharged frequently. Based on this, the capacitance value is optimized from 1000μF to 1200μF to improve the response speed of the topology during the simulation process.

[0029] In another embodiment of the present invention, the voltage waveform of a 10kV arm under 50Hz switching is simulated in a simulation environment. In the initial configuration, the waveform may have an overshoot of 5%. After adjustment through the characteristic distribution, the overshoot is reduced to 2%, indicating that the parameter optimization is effective. If the calculation resource occupancy exceeds the standard, such as the CPU usage rate reaches 90% exceeding the threshold of 80%, then through the load balancing algorithm, the simulation tasks are allocated to a multi-core processor, and the usage rate is reduced to 70% after optimization, improving the simulation efficiency.

[0030] S2. Perform a fault simulation on the to-be-tested symmetrical single-pole de-icing device with preset pole-to-ground fault parameters to obtain the voltage transient waveforms of the sub-modules of the to-be-tested symmetrical single-pole de-icing device under different switching states.

[0031] After the simulation environment is set up, the pole-to-ground fault parameters will be defined to simulate the pole-to-ground faults that may occur during the actual operation of the symmetrical single-pole de-icing device. Specifically, the pole-to-ground faults include fault types such as single-phase grounding, two-phase grounding, three-phase grounding, and pole-to-ground short circuit, and the fault locations involve the upper arm, lower arm, or specific sub-modules of the arm.

[0032] The fault parameters to be set include the transition resistance value, phase current / voltage value, short-circuit resistance value, and duration, etc. Taking the pole-to-ground short-circuit fault as an example: Suppose the short-circuit fault occurs in the 3rd sub-module of the upper arm, and the fault parameters can be set such as the short-circuit resistance is 0.1 ohm and the fault triggering time is 0.5 seconds. For the above short-circuit fault, the simulation result will show that the voltage of the 3rd sub-module instantaneously rises to 1500 volts (the normal value is 1000 volts).

[0033] It should be understood that during the simulation process, the pre-configured sub-module switching strategy, that is, the input, cut-off, and switching frequencies of the sub-modules, will also be involved in the simulation. Specifically, in different pole-to-ground fault scenarios, by injecting different input / cut-off states, the fault scenarios are run in the electromagnetic transient simulation environment to collect the port voltage transient waveform data of each sub-module. Preferably, the sampling rate is set ≥10kHz.

[0034] S3. Conduct voltage stress analysis on the voltage transient waveform to obtain the voltage stress distribution data and waveform anomaly points of the sub-module.

[0035] It can be understood that the voltage transient waveform in the initial stage of the fault contains core information such as the fault type (short circuit / grounding) and location (upper / middle section of the arm), which is the data basis for subsequent device diagnosis and control. That is, the voltage transient waveform records the information of the voltage change over time during the fault process, including the peak voltage, rising rate, and duration, and these indicators reflect the voltage stress distribution of the sub-module in the initial stage of the fault.

[0036] Among them, the peak voltage refers to the maximum value of the instantaneous voltage, which reflects the overvoltage risk; the rising rate refers to the speed at which the voltage rises from the steady state to the peak value, reflecting the rapid release of electromagnetic energy; the duration refers to the time when the voltage exceeds the safety threshold, which is related to the tolerance ability of the equipment.

[0037] Specifically, in the embodiment of the present invention, the voltage signal data in the voltage transient waveform is extracted, and the input and cut-off states of the sub-module are divided according to the voltage signal data. Exemplarily, the voltage signal data in the voltage transient waveform can be extracted by a high-speed data collector, or the waveform can be subjected to spectrum analysis by using Fourier transform, and the noise is filtered through filter processing to obtain the required voltage signal data (i.e., voltage amplitude, time, rising / falling rate, oscillation frequency, etc.).

[0038] Regarding the process of dividing the sub-module switching state in the waveform, in this embodiment, traversal is performed in the waveform by setting a division threshold. Specifically, the difference between the voltage signal data and the preset rated voltage value of the sub-module is calculated, and the division threshold is set based on the difference. Based on the division threshold, the sliding window algorithm is used to traverse in the voltage transient waveform to determine the switching state of the sub-module. Exemplarily, if the determination threshold for the input state is set to 1.1 times the rated value and lasts for more than 1 ms, and the determination threshold for the cut-off state is set to 0.9 times the rated value.

[0039] It can be understood that the switching of the switching state directly affects the voltage distribution. When a sub-module is input or cut off, the charging and discharging process of the capacitor will change its voltage. If the switching is unbalanced, some sub-modules may charge and discharge frequently, resulting in large fluctuations in the capacitor voltage, while the voltage of other sub-modules remains stable. This will affect the voltage distribution of the entire converter, may cause voltage imbalance, increase harmonics, and even damage the equipment.

[0040] Based on this, in this embodiment, according to the partitioning result, i.e., the switched states partitioned, the voltage stress index values corresponding to the sub-modules in each switched state are calculated. That is, the quantization results of the peak voltage, the rising rate, and the duration. Exemplarily, as shown in the simulation process, the voltage of the sub-module in the bypass state is 0V, and after being switched in, the detected voltage peak is 1800V, exceeding the safety threshold of 1500V. By comparing the data at different times, it can be found that the voltage stress increases significantly at the moment of switching in, while it tends to be stable during bypass, revealing the law of state switching and voltage stress change.

[0041] Furthermore, statistical analysis is performed on the voltage stress index values to obtain voltage stress distribution data. In this embodiment, multi-index joint analysis can be performed on the voltage stress index values to fit the voltage stress distribution curve of the sub-module, and the voltage stress distribution can be visually obtained from the curve. Exemplarily, three indexes, namely the peak voltage, the rising rate, and the duration, can be combined to draw a two-dimensional histogram or a contour map to obtain the distribution curve.

[0042] Furthermore, abnormal waveform segments can be marked according to these index values, and waveform abnormal points in the voltage transient waveform can be marked according to the screening factor calculated from the voltage stress index values. In this embodiment, it should be understood that the abnormal segment refers to a specific time interval marked in the voltage transient waveform due to overvoltage or the voltage change rate exceeding the preset threshold, which can be used for the accurate positioning of the root cause of equipment failure in the subsequent process, and can also quantify the tolerance ability of the equipment.

[0043] Among them, the screening factor is used to screen out abnormal wave bands that exceed / fall below the set threshold in the waveform. Specifically, the overvoltage multiple and the voltage change rate are calculated according to the voltage stress index values, and the overvoltage multiple and the voltage change rate are used as double screening factors. When any one of the double screening factors exceeds the preset abnormal threshold, waveform abnormal points will be marked in the voltage transient waveform. Exemplarily, the overvoltage multiple is determined by the ratio between the peak voltage and the rated voltage. If the peak voltage is 2kV and the rated voltage is 1kV, then the overvoltage multiple is 2 / 1 = 2 p.u. And if the set overvoltage multiple threshold is 1.5kV, it indicates that the voltage stress exceeds the safety range and needs to be marked as an abnormal segment.

[0044] The calculation of the voltage change rate (dv / dt) requires the voltage difference and the time difference. In this embodiment, the voltage-time series is extracted from the waveform, and the differential method is used to calculate the instantaneous change rate. Exemplarily, the ratio of the difference between the voltage values corresponding to the 1.1ms moment and the 1.0ms moment to the time change amount is calculated to obtain the corresponding voltage change rate. When this change amount exceeds the threshold (such as 7kV / ms), it indicates that the voltage stress exceeds the safety range and needs to be marked as an abnormal segment.

[0045] The above process takes the switching state of the sub-module as an index to evaluate the voltage stress distribution borne by sub-modules in different switching states. In some embodiments of the present invention, the topology structure of the sub-module is further introduced as an evaluation index to further analyze the voltage stress distribution of the sub-module under different topological connection modes (such as half-bridge, full-bridge or hybrid structure). It should be understood that the number and connection mode of switching devices in different topological structures are different, which will affect the voltage distribution path and the voltage withstand capacity of the devices. Moreover, in the case of a fault, the fault current blocking capabilities of different topological structures are different, which will also affect the voltage stress distribution during the fault.

[0046] Based on this, in this embodiment, the topological connection mode of the sub-module in the symmetric monopole ice melting device to be tested is adjusted. For example, it is adjusted from a full-bridge structure to a half-bridge structure, and the number of branches is correspondingly adjusted, and the simulation environment is updated with the adjusted first topological structure.

[0047] In the updated simulation environment, the input / removal logic of the sub-module is controlled to perform fault simulation again, and the corresponding comparative voltage transient waveforms of the sub-module under different topological structures are obtained according to the simulation results. It should be noted that the switching logic is usually determined by the pulse width modulation signal. For example, adjusting the modulation ratio from 0.8 to 0.9 may cause the switching frequency of the sub-module to increase, and the voltage stress to rise from 1600V to 1900V. Thus, it can also be judged that the switching control strategy needs to balance the switching frequency and the stress level to avoid frequent switching and amplifying the stress.

[0048] Similar to the above analysis method, statistical analysis is performed on the comparative voltage transient waveforms, the voltage stress index values are calculated, and the comparative voltage stress distribution data is obtained.

[0049] Exemplarily, it is assumed that the sub-module adopts a half-bridge topology. When in the input state, the capacitor bears all the voltage, and there is no stress when bypassed. In an arm containing 10 sub-modules, a certain sub-module has a higher voltage stress of 2000V because its topological position is close to the midpoint of the arm compared with the two ends. And this distribution difference actually also indicates that the topological design of this device needs to be optimized.

[0050] Regarding the influence of the above switching state and topological structure on the voltage stress distribution, this embodiment further analyzes the importance of the two influences. Specifically, in this embodiment, the voltage stress distribution data and the comparative voltage stress distribution data are compared and analyzed to determine the high-order index affecting the voltage stress of the sub-module.

[0051] Specifically, first, preprocess the first topology structure, the switching states of sub-modules, the voltage stress distribution data, and the comparative voltage stress distribution data to obtain the dataset to be input. Exemplarily, for a changed topology structure, each of its structure types can be mapped into a one-hot encoded form, and the number of branches can be standardized. For the switching states, they can be discretized.

[0052] Input the dataset to be input into the trained random forest model, output the importance scores affecting the voltage stress index, sort the importance scores, and determine the high-order indexes that have a greater impact on the voltage stress distribution according to the sorting results. For example, if the random forest model output shows that the importance of the topology type is 28% and the switching state is 32%, then the switching state is the high-order influencing index. Exemplarily, in the subsequent device optimization process, more attention will be paid to formulating a more reasonable switching control strategy.

[0053] Furthermore, in this embodiment, the simulation process will be optimized according to the differences in the voltage stress distribution under different index variables to improve the accuracy of the simulation results.

[0054] Specifically, first, according to the voltage stress distribution data, compare the first voltage stress differences of the sub-modules under different switching states. Exemplarily, the voltage stress distribution curves under different switching states previously fitted can be compared and analyzed. Or compare the mean differences between the voltage stress index value data of two groups under different switching states.

[0055] Similarly, according to the comparative voltage stress distribution data, use methods such as the mean calculation method and the principal component analysis method to compare the second voltage stress differences of the sub-modules under different topological connection modes.

[0056] Input the preprocessed first voltage stress difference and second voltage stress difference into the pre-trained difference classification model to obtain the stress difference levels of each sub-module. For example, the output labels are 0 / 1 / 2, corresponding to the results of low / medium / high risk respectively. For example, these difference data can be cleaned and standardized. Preferably, in this embodiment, a support vector machine pre-trained difference classification model is used.

[0057] Modify the sub-module topology structure and switching state parameters according to the stress difference levels, and re-iterate the simulation and stress analysis processes. Exemplarily, if the output is a high-risk difference, the switching strategy can be optimized to reduce the switching frequency. If the output is a medium risk, the topology structure can be adjusted, and the half-bridge can be replaced with a hybrid bridge. If the output is a low risk, the current parameters can be maintained and the current simulation test can be continued.

[0058] S4 - S5: Input the voltage stress distribution data and waveform anomaly points into the pre - trained fault prediction model for processing to obtain the fault prediction results corresponding to the sub - modules. Based on the fault prediction results, determine the fault risk level of the symmetric monopolar ice - melting device to be measured.

[0059] This step is the process of fault prediction. Preferably, a support vector machine model is trained to obtain this fault prediction model. The model will output the classification probability of fault types, the confidence interval of fault locations, and a list of abnormal sub - module numbers, so as to accurately locate the fault location and identify the fault type. In some embodiments of the present invention, when using the support vector machine algorithm to establish a fault prediction model, it can be trained based on key input variables. For example, if the input data set contains two types of samples, normal and faulty, the voltage peak of normal samples is mostly below 580V, while that of faulty samples is above 620V. The model determines the fault type such as "over - voltage fault" and the location such as "the 3rd sub - module of the arm" by finding the optimal classification boundary.

[0060] After obtaining the above - mentioned fault prediction results, this embodiment will further conduct a risk assessment on the fault prediction results. Extract features from the fault prediction results to obtain a fault feature data set. These features include but are not limited to: voltage peak, ripple coefficient, current amplitude, and fault type, etc.

[0061] Input the fault feature data set into the pre - trained risk prediction model to obtain the fault risk level, and formulate corresponding maintenance strategies for the symmetric monopolar ice - melting device to be measured based on the fault risk level. Exemplarily, a risk prediction model can be trained by using a random forest model, a LightGBM gradient - boosting decision tree model, etc.

[0062] Exemplarily, if the level is high - risk: predicted as a three - phase ground fault, and the anomaly point density > 5 points / ms, or the peak mean exceeds the threshold by 30%, it is recommended to immediately cut off the faulty arm and enable the maintenance strategy of the redundant sub - module.

[0063] If the level is low - risk: predicted as a single - phase ground fault, the anomaly point density < 2 points / ms, and the peak mean does not exceed the standard, it is recommended to conduct regular inspections and record the trend of anomaly points.

[0064] In summary, in the embodiments of the present invention, an electromagnetic transient simulation environment is constructed to analyze the voltage transient characteristics under pole-to-ground faults. Combining the switching states of sub-modules, the voltage peak value, rising rate, and duration for evaluating the voltage stress distribution are analyzed, and the abnormal points are located. Further, a machine learning model is established using a support vector machine. With the voltage stress distribution and abnormal points as inputs, the fault types and locations occurring in the ice melting device are predicted. It can quickly distinguish fault characteristics, effectively locate the source of the problem, and finally evaluate the damage risk level of sub-modules based on the prediction results, so as to formulate effective maintenance and early warning strategies. The present invention effectively integrates dynamic simulation and machine learning technologies, makes up for the deficiencies of traditional steady-state analysis, and provides an accurate risk assessment basis for improving equipment reliability and power grid safe operation.

[0065] An embodiment of the present invention provides a system. Specifically, please refer to Figure 2 , Figure 2 which is shown as the structural schematic diagram of the pole-to-ground fault analysis system of the ice melting device considering transient characteristics in one of the embodiments of the present invention, including the following: A simulation initialization module M1, configured to create a simulation environment according to the topological structure of the to-be-tested symmetric single-pole ice melting device and the historical operation data of sub-modules; A fault simulation module M2, configured to perform fault simulation on the to-be-tested symmetric single-pole ice melting device with preset pole-to-ground fault parameters to obtain the voltage transient waveforms of the sub-modules of the to-be-tested symmetric single-pole ice melting device under different switching states; A transient analysis module M3, configured to perform voltage stress analysis on the voltage transient waveforms to obtain the voltage stress distribution data and waveform abnormal points of the sub-modules; A fault prediction module M4, configured to input the voltage stress distribution data and the waveform abnormal points into a pre-trained fault prediction model for processing to obtain the fault prediction results corresponding to the sub-modules; A risk analysis module M5, configured to determine the fault risk level of the to-be-tested symmetric single-pole ice melting device according to the fault prediction results.

[0066] The technical features and technical effects of the system proposed in the embodiments of the present invention are the same as those of the method proposed in the embodiments of the present invention, and will not be elaborated here.

[0067] The above embodiments only represent several implementation manners of the present invention. The description is relatively specific and detailed, but it cannot be understood as a limitation to the scope of the patent of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several deformations and improvements can still be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention should be subject to the appended claims.

Claims

1. A method for analyzing the pole-to-ground fault of an ice melting device considering transient characteristics, characterized in that Including: Create a simulation environment according to the topological structure of the symmetric monopole ice melting device to be tested and the historical sub-module operation data; Perform fault simulation on the symmetric monopole ice melting device to be tested with preset pole-to-ground fault parameters to obtain the voltage transient waveforms of the sub-modules of the symmetric monopole ice melting device to be tested in different switching states; Conduct voltage stress analysis on the voltage transient waveforms to obtain the voltage stress distribution data and waveform anomaly points of the sub-modules; Input the voltage stress distribution data and the waveform anomaly points into a pre-trained fault prediction model for processing to obtain the fault prediction results corresponding to the sub-modules; Determine the fault risk level of the symmetric monopole ice melting device to be tested according to the fault prediction results.

2. The method for analyzing the pole-to-ground fault of the ice melting device considering transient characteristics according to claim 1, characterized in that Before performing the fault simulation on the symmetric monopole ice melting device to be tested with preset pole-to-ground fault parameters, it further includes: Extract features from the historical switching data in the historical sub-module operation data to construct a switching feature data set; Input the switching feature data set into a pre-constructed statistical distribution model and output the feature distribution during the sub-module switching process; Adjust the sub-module parameters of the symmetric monopole ice melting device to be tested according to the feature distribution, and update the simulation environment with the adjusted sub-module parameters and the topological structure.

3. The method for analyzing the pole-to-earth fault of the ice melting device considering transient characteristics according to claim 1, characterized in that, The process of conducting voltage stress analysis on the voltage transient waveforms to obtain the voltage stress distribution data and waveform anomaly points of the sub-modules includes: Extract the voltage signal data in the voltage transient waveforms and divide the input and cut-off states of the sub-modules according to the voltage signal data; Calculate the voltage stress index values in each switching state respectively according to the division results; Conduct statistical analysis on the voltage stress index values to obtain the voltage stress distribution data, and mark the waveform anomaly points in the voltage transient waveforms according to the screening factors calculated from the voltage stress index values.

4. The method for analyzing the pole-to-ground fault of the ice melting device considering transient characteristics according to claim 3, characterized in that, The process of conducting statistical analysis on the voltage stress index values to obtain the voltage stress distribution data, and marking the waveform anomaly points in the voltage transient waveforms according to the screening factors calculated from the voltage stress index values includes: Conduct multi-index joint analysis on the voltage stress index values to fit the voltage stress distribution curve of the sub-module; and, Calculate the overvoltage multiple and voltage change rate according to the voltage stress index values, and use the overvoltage multiple and the voltage change rate as double screening factors; When any index in the double screening factors exceeds the preset anomaly threshold, mark the waveform anomaly points in the voltage transient waveforms.

5. The method for analyzing the pole-to-ground fault of the ice melting device considering transient characteristics according to claim 3, wherein The process of dividing the input and cut-off states of the sub-modules according to the voltage signal data includes: Calculate the difference between the voltage signal data and the preset rated voltage value of the sub-module, and set the division threshold with the difference; Traverse in the voltage transient waveforms using the sliding window algorithm based on the division threshold to determine the switching states of the sub-modules.

6. The method for analyzing the pole-to-ground fault of the ice melting device considering transient characteristics according to claim 3, wherein The process of conducting voltage stress analysis on the voltage transient waveforms further includes: Adjust the topological connection mode of the sub-modules in the symmetric monopole ice melting device to be tested, and update the simulation environment with the adjusted first topological structure; In the updated simulation environment, the input / removal logic of the control sub-module conducts fault simulation again; Based on the simulation results, obtain the comparative voltage transient waveforms corresponding to the sub-modules under different topological structures; Determine the comparative voltage stress distribution data according to the comparative voltage transient waveforms; Compare and analyze the comparative voltage stress distribution data and the voltage stress distribution data to determine the high-order indicators affecting the sub-module voltage stress.

7. The method for analyzing the pole-to-earth fault of the ice melting device considering transient characteristics according to claim 6, characterized in that, The step of comparing and analyzing the comparative voltage stress distribution data and the voltage stress distribution data to determine the high-order indicators affecting the sub-module voltage stress includes: Preprocess the first topological structure, sub-module switching state, the voltage stress distribution data, and the comparative voltage stress distribution data to obtain the dataset to be input; Input the dataset to be input into the trained random forest model and output the importance scores of the indicators affecting the voltage stress; Sort the importance scores and determine the high-order indicators according to the sorting results.

8. The method for analyzing the pole-to-ground fault of the ice melting device considering transient characteristics according to claim 6, characterized in that The process of conducting voltage stress analysis on the voltage transient waveforms further includes: According to the voltage stress distribution data, compare the first voltage stress differences of the sub-modules under different switching states; According to the comparative voltage stress distribution data, compare the second voltage stress differences of the sub-modules under different topological connection modes; Input the preprocessed first voltage stress difference and the second voltage stress difference into the pre-trained difference classification model to obtain the stress difference levels of each sub-module; Modify the sub-module topological structure and switching state parameters according to the stress difference levels and re-iterate the simulation analysis process.

9. The method for analyzing the pole-to-ground fault of the ice melting device considering transient characteristics according to claim 1, wherein The step of determining the fault risk level of the to-be-tested symmetric monopole ice melting device according to the fault prediction result includes: Extract features from the fault prediction result to obtain a fault feature dataset; Input the fault feature dataset into the pre-trained risk prediction model to obtain the fault risk level, and formulate a corresponding maintenance strategy for the to-be-tested symmetric monopole ice melting device based on the fault risk level.

10. An ice melting device pole-to-earth fault analysis system considering transient characteristics, characterized in that, It includes: A simulation initialization module for creating a simulation environment according to the topological structure of the to-be-tested symmetric monopole ice melting device and the historical sub-module operation data; A fault simulation module for conducting fault simulation on the to-be-tested symmetric monopole ice melting device with preset pole-to-earth fault parameters to obtain the voltage transient waveforms of the sub-modules of the to-be-tested symmetric monopole ice melting device under different switching states; A transient analysis module for conducting voltage stress analysis on the voltage transient waveforms to obtain the voltage stress distribution data and waveform abnormal points of the sub-modules; A fault prediction module for inputting the voltage stress distribution data and the waveform abnormal points into the pre-trained fault prediction model for processing to obtain the fault prediction results corresponding to the sub-modules; A risk analysis module for determining the fault risk level of the to-be-tested symmetric monopole ice melting device according to the fault prediction result.

Images